Log-logistic flood frequency analysis

doi:10.1016/0022-1694(88)90015-7

Journal of Hydrology

Volume 98, Issues 3–4, 15 April 1988, Pages 205-224

https://doi.org/10.1016/0022-1694(88)90015-7 Get rights and content

Abstract

The log-logistic (LLG) distribution is evaluated for flood frequency analysis. Some of its properties and methods of parameter estimation are given, including a new method based on generalised least squares (GLS). The performance of the log-logistic distribution is compared with those of the generalised extreme value (GEV), three parameter log-normal (LN3) and three parameter Pearson (P3) distributions. The results are reported in terms of empirical distribution function (EDF) tests of goodness of fit, on both individual and regional flood series through the application of these distributions to a set of reasonably long annual maximum series for part of Scotland. Some reproductive properties of the LLG and GEV are also compared. In terms of four key properties the LLG performs better than the GEV, LN3 and P3 distributions and is thus commended for further analysis.

References (24)

M.C. Acreman et al.
Classification of drainage basins according to their physical characteristics; an application for flood frequency analysis in Scotland
J. Hydrol.
(1986)
R.F. Kappenman
Estimation for the three parameter Weibull, Lognormal and Gamma distributions
Comput. Stat. Data Anal.
(1985)
M.C. Acreman
Estimating flood statistics from basin characteristics in Scotland
C. Cunnane
Review of statistical models for flood frequency estimation
GENSTAT
Optimisation
J.A. Greenwood et al.
Probability weighted moments
Definition and relation to parameters of distributions expressible in inverse form
Water Resour. Res.
(1979)
J.R.M. Hosking
The theory of probability weighted moments
J.R.M. Hosking et al.
An appraisal of the regional flood frequency procedure in the U.K. Flood Studies Report
Hydrol. Sci. J.
(1985)
J.R.M. Hosking et al.
Estimation of the Generalised Extreme Value distribution by the method of probability weighted moments
Technometrics
(1985)
J.C. Houghton
Birth of a parent: The Wakeby distribution for modelling flood flows
Water Resour. Res.
(1978)

J.F. Lawless

A note on lifetime regression models

Biometrika

(1986)

J.C. Matalas et al.

Eureka! It fits a Pearson type 3 distribution

Water Resour. Res.

(1973)

Cited by (195)

Intelligent optimal preventive replacement maintenance policy for non-repairable systems
2024, Computers and Industrial Engineering
This paper developed optimal maintenance policies for industrial machines to ensure uninterrupted production. Analytical and intelligent replacement models were constructed using real-time data from a commercial photocopier operator’s expenditure logbook between 2012 and 2023. A goodness-of-fit test affirmed a log-logistic failure distribution with shape and scale parameters of 1.7233 and 763.9220, respectively. Probability functions from this distribution established replacement models, determining optimal preventive maintenance conditions. Through feature engineering additional features crucial for operational and cost optimization were abstracted, with 4,881 unique data points simulated from boundary conditions defined by the original or raw data. Analytical results showed that the failure distribution-based replacement model outperformed the hazard function-based model in minimum replacement time and unit maintenance cost over time, while key parameters such as availability, reliability, and failure occurrence probability remained constant at 96 %, 94 %, and 0.07 %, respectively. Machine learning models (Extreme Gradient Boosting: XGBoost, Support Vector Machine: SVM, and Multilayer Perceptron: MLP) demonstrated very high accuracy in availability (XGBoost = 99.49 %, SVM = 99.18 %, RF = 100 %, MLP = 99.80 %), reliability (XGBoost = 99.49 %, SVM = 99.18 %, RF = 100 %, MLP = 99.80 %), and maintenance cost (XGBoost = 100 %, SVM = 100 %, RF = 100 %, MLP = 100 %), exceeding results of the analytical models.
Reduced bias estimation of the shape parameter of the log-logistic distribution
2024, Journal of Computational and Applied Mathematics
In the literature, the log-logistic distribution is commonly presented with two parameters: one that governs the shape of the model, and the other that governs its scale. However, to make this model more suitable for data analysis, an additional location parameter can be added, resulting in the three-parameter or shifted log-logistic model. In this paper, we introduce a new estimator for the shape parameter of a three-parameter log-logistic distribution that reduces bias. We also derive various properties of the proposed estimator. Additionally, a simulation study and an application example to a real data set are conducted to examine the efficiency for finite sample sizes. The theoretical and simulated results confirm that our proposed estimation method performs significantly better than other estimation methods found in the literature.
A new alpha logarithmic-generated class to model precipitation data with theory and inference
2023, Heliyon
Precipitation, or rainfall, is a central feature of the weather cycle and plays a vital role in sustaining life on Earth. However, existing models such as the Poisson, exponential, normal, log-normal, generalized Pareto, gamma, generalized extreme value, lognormal, beta, Gumbel, Weibull, and Pearson type III distributions used for predicting precipitation are often inadequate for precisely representing the complex pattern of rainfall. This study proposes a novel and innovative approach to address these limitations through the new alpha logarithmic-generated (NAL-G) class of distributions. The study authors thoroughly examine the NAL-G class and a unique model, the NAL-Exponential (NAL-Exp) distribution, with a focus on analyzing mathematical properties such as moments, quantile function, entropy, order statistics, and more. Six recognized classical estimation methods are employed, and extensive simulations are conducted to determine the best one. The NAL-Exp distribution demonstrates its high adaptability and value through its superior performance in modeling four distinct rainfall data sets. The results show that the NAL-Exp distribution outperforms other commonly used distribution models, highlighting its potential as a valuable tool in hydrological modeling and analysis. The increased versatility and flexibility of this new approach hold great potential for enhancing the accuracy and reliability of future rainfall predictions.
Spectral wave energy period and peak period statistics concomitant with maximum significant wave heights
2023, Coastal Engineering
For this investigation, we employ maximum significant wave heights associated spectral wave energy periods off Aguçadoura (41.46°N, −8.85°W), off Portugal in the North Atlantic Ocean, which were generated over a period of 34 years (1979–2012) by the wave model (SWAN). Conditional Mean Functions of generalised Pareto, Erlang, and three-parameter Weibull distributions are taken into account to determine the distribution of wave energy period. The values computed from the SWAN model data are reasonably coherent with the mean, mean maximum, and most frequent maximum energy periods, as well as the average of the one-third and one-tenth of the highest energy periods projected from the models. The generalised Pareto distribution is used to develop a parametric relation (based on order statistics) that can be used to accurately predict the average of one-third of the highest wave periods. Substantial linear functional relationships exist between computed besides estimated energy period versus peak period statistics. The average of one-third of the highest energy periods to mean energy periods accurately depicts the empirical ratio of significant wave period to mean wave period of 1.2. Daily maximum significant wave heights and associated spectral wave energy periods and peak periods by WAMAN (Wave monitoring along nearshore) wave rider buoy off Kozhicode (11.2698°N, 75.6722°E), southwest coast of India in the Arabian Sea during the winter season (January 2018 to April 2018) are also subjected to Conditional Mean Function Analysis. The sample values are consistent with the population estimates of the mean and the average of the one-third of the highest wave periods. An attempt is made to approximate the range of peak periods associated with extreme significant wave heights by taking into account the steepness constraints of a wave profile and dispersion relations for spectral peak periods.
Bivariate socioeconomic drought assessment based on a hybrid framework and impact of human activities
2023, Journal of Cleaner Production
Socioeconomic drought generally occurs when water supply cannot satisfy water demand. This study proposed a hybrid framework for socioeconomic drought assessment, involving the construction of standardized water supply-demand reliability index (SWSDRI), truncation of drought events, and classification of bivariate drought levels. Using the run theory, socioeconomic drought events and their properties, including duration and severity, were characterized by the SWSDRI series. Then, bivariate socioeconomic drought levels, based on the Kendall frequency of drought duration and severity, were classified using the copula function. Impacts of human activities on socioeconomic drought, including mainly the regulation of reservoirs and water consumption were quantified in the Dongjiang River basin, located in the south of China. Results showed that the framework, based on SWSDRI, facilitated the investigation of socioeconomic drought properties. The joint distributions of drought duration and severity, based on naturalized streamflow and under the individual and total impacts of human activities, were fitted well by the Archimedean copulas. Both drought duration and severity clearly exhibited the slight, moderate, severe, and extreme droughts classified by bivariate drought levels. For naturalized streamflow, socioeconomic drought occurred approximately once a year in the Dongjiang River basin. Design values of the combination of drought duration and severity accounted for 3.4 months and 8.0 hundred million cubic meters of the boundary between slight and moderate droughts, 4.9 months and 13.6 hundred million cubic meters of the boundary between moderate and severe droughts, and 6.9 months and 20.0 hundred million cubic meters of the boundary between moderate and severe droughts, respectively. Under the impact of human activities, in particular the regulation of reservoirs, socioeconomic drought was remarkably alleviated with a decrease of 45.8% in quantity. If the impact of water consumption was considered, drought events in quantitive terms remained almost unchanged, but more extreme and severe droughts occurred, because some slight, moderate, and severe droughts were aggravated by one level. Among the main reservoirs, the impact of the Xinfengjiang reservoir in the period of its operation was remarkable with a decrease in drought events by 47.7%. Under the impact of the Fengshuba reservoirs, drought events changed little in quantitive terms, but more slight droughts occurred because some moderate, severe, and extreme droughts were alleviated.
Multi-steps drought forecasting in arid and humid climate environments: Development of integrative machine learning model
2023, Agricultural Water Management
In the current study, the Standardized Precipitation Evaporation Index (SPEI) was forecasted using a combination of the empirical wavelet transform (EWT), discrete wavelet transforms (DWT), extended Kalman filter (EKF), two models of multilayer perceptrons (MLP), and group method of data handling (GMDH) neural networks. Two synoptic stations of Tabriz (semi-arid climate) and Rasht (humid climate) covering data period (1987–2019) were selected for forecasting. 70% of the data was used for model training and 30% for validation. Three forecasting horizons (1, 3, and 6 months ahead SPEI) were investigated. Autocorrelation function and partial autocorrelation function were used to determine the optimal inputs to the models. The outcomes of the present study showed that in both stations, the combination of machine learning models with two types of wavelet transforms (EWT and DWT) compared to the standalone models improved the performance of the forecasting (correlation coefficient, R = 0.9980, root mean square error, RMSE = 0.0483 for Tabriz Station and R = 0.9988, RMSE = 0.0521 for Rasht Station). A comparison of the EWT and DWT wavelets showed that the EWT had better performance in all forecasting intervals. By raising the forecasting interval from one month to six months, EWT performance was more evident than DWT performance. In 6-month forecasting horizon, the DWT had almost no effect on model performance improvement. In both stations, the combination of the EWT and MLP-EKF model had the best performance in forecasting SPEI drought index.

View all citing articles on Scopus

View full text

Research paperLog-logistic flood frequency analysis

Abstract

J. Hydrol.

Comput. Stat. Data Anal.

Estimating flood statistics from basin characteristics in Scotland

Review of statistical models for flood frequency estimation

Optimisation

Probability weighted moments

Definition and relation to parameters of distributions expressible in inverse form

Water Resour. Res.

The theory of probability weighted moments

An appraisal of the regional flood frequency procedure in the U.K. Flood Studies Report

Hydrol. Sci. J.

Estimation of the Generalised Extreme Value distribution by the method of probability weighted moments

Technometrics

Birth of a parent: The Wakeby distribution for modelling flood flows

Water Resour. Res.

A note on lifetime regression models

Biometrika

Eureka! It fits a Pearson type 3 distribution

Water Resour. Res.

Research paper
Log-logistic flood frequency analysis